A free-energy approach for stochastic reaction-drift-diffusion lattice simulations of crowded biological systems
Molecular size alters spatial distributions, reaction equilibria, and transport in crowded mixtures, but resolving individual particle encounters can make large stochastic simulations computationally costly. We developed a lattice-based framework for simulating reaction-drift-diffusion systems with volume exclusion captured as a free energy functional in each voxel. Changes in this free energy, together with external one-body potentials, modify diffusion and reaction propensities. Our framework accommodates mixtures of particle sizes without resolving each particle's position within a voxel. We implemented our framework in an operator splitting scheme and compared our predictions with off-lattice interacting-particle simulations. Inclusion of the local free energy functional improved spatial occupancies and shifts in reaction equilibria relative to point-particle simulations. With separate approximations for larger spheres and polymers, we simulated crowding-induced slowing and spatial segregation. We then modeled motor proteins' traffic jams on a structured track representing a microtubule. Our results demonstrate how local composition can inform stochastic reaction and transport rates without resolving individual encounters.